RESEARCH
All researchLouder won't make you stand out. Different will.
Different is seen, similar has to be searched for. So the difference is chosen before the look: two strategic moves come before the visual one.
added to the search by each extra letter on screen when the target had its own colour or shape.
TREISMAN AND GELADE, 1980 · 6 PEOPLE, 1 TO 30 LETTERSLIMIT · A LAB TEST WITH 6 PEOPLE AND LETTERS, NOT BUYERS. IT SHOWS HOW THE EYE WORKS.The question.
Every black suitcase at the airport looks the same, so people tie a ribbon on theirs. To a buyer comparing suppliers, a company whose product is years ahead can still be one more black suitcase. The usual answer is to get louder. The research on how people find things points somewhere else: what gets found is what is different, so the question is what the difference is and in which order to build it.
How we know.
Treisman and Gelade published the search experiment in Cognitive Psychology in 1980. Six people from the Oxford Subject Panel, aged 24 to 29, looked for a target among 1, 5, 15 or 30 coloured letters while their response time was measured. In one condition the target had a colour or a shape nothing else had. In the other it shared its colour with some letters and its shape with the rest. The buyer figures come from two separate surveys: 9,000 respondents in CEB's 2009 Customer Experience Survey, and brand users in 17 consumer markets studied by Romaniuk, Sharp and Ehrenberg in 2007.
What it means.
A shortlist is a display too. If your homepage, deck and stand share their colours, images and promises with the rest of the category, a buyer has to search for you instead of seeing you. But the experiment is about finding, not buying, so a new look only earns its place when it carries a difference you can prove. That is why two strategic moves come before the visual one.
What each extra letter cost.
MILLISECONDS ADDED TO THE SEARCH BY EACH EXTRA LETTER ON SCREENTREISMAN AND GELADE, 1980, EXPERIMENT I · 6 PEOPLE. FITTED SLOPES FOR TRIALS WHERE THE TARGET WAS PRESENT. BARS ARE DRAWN AGAINST THE LARGER FIGURE.
Three pieces of evidence.
ONE EXPERIMENT ON SEEING, TWO SURVEYS ON SAMENESS28.7 milliseconds per letter against 3.1, from the paper's fitted slopes for trials where the target was there to be found.
The other 86% did not, which is the sameness a new supplier walks into.
Being told apart on meaning is rare even among customers, so pick a difference a stranger can check.
How few people see a difference.
SHARE OF A BRAND'S OWN USERS WHO CALLED IT DIFFERENT OR UNIQUEROMANIUK, SHARP AND EHRENBERG, 2007. CONSUMER BRANDS IN AUSTRALIA AND THE UK, NOT BUSINESS BUYERS. THE EXCEPTION IS A STORE THAT DOES NOT STOCK NATIONAL BRANDS: A DIFFERENCE ANYONE CAN CHECK.
In depth.
THE LONG READWhat Treisman and Gelade measured in 1980.
In 1980 Treisman and Gelade published a search experiment in Cognitive Psychology. Six people from the Oxford Subject Panel, aged 24 to 29, looked at cards carrying 1, 5, 15 or 30 coloured letters and pressed a key when they found a target. The other letters were brown Ts and green Xs. In one condition the target had a feature nothing else had: it was a blue letter or an S. In the other it was a green T, which shared its colour with the Xs and its shape with the Ts. When the target had its own colour or shape, each extra letter on the card added 3.1 milliseconds to the search. When it shared its features, each extra letter added 28.7. With a unique feature, the authors report, the number of other letters hardly changed the search time.
What the two slopes add up to.
The two figures are slopes, fitted by the authors to the trials where the target was present. Put through the paper's own equations, a search of 30 letters takes about 0.54 seconds when the target is unique and about 1.26 seconds when it shares its features. Per extra letter, the similar target costs 9.3 times more. That ratio is our arithmetic on the paper's numbers, not a figure the paper prints. A unique target is found at close to the same speed among 5 letters or 30. A target built from the same parts as its neighbours gets slower with every neighbour added. Each person did three blocks of 128 trials in each condition, the first block was treated as practice, and all six people showed the same pattern.
What it does not show.
This is a laboratory test with six people and letters, not buyers and suppliers. It measures how quickly the eye finds a target. It says nothing about preference, trust or price. Later research also softened the clean split. In 1989 Duncan and Humphreys argued in Psychological Review that search is not two types but one scale: it gets harder as the target becomes more similar to the other items, and as the other items become less similar to each other. In 1998 Wolfe, writing in Psychological Science, looked back at about 1 million trials from 2,500 sessions and found one continuous spread of slopes with no line dividing easy searches from hard ones. So different is seen is a matter of degree. Even the 3.1 figure is a summary: the paper reports that the unique-target results departed from a straight line.
How alike suppliers look to the people choosing.
Two surveys suggest sameness is the normal condition. CEB's 2009 Customer Experience Survey asked 9,000 respondents whether they saw a real difference between suppliers and valued it enough to pay for it. 14% said yes and 86% said no. The figure is printed in From Promotion to Emotion, a 2013 paper by CEB with Google, which argues that business value gets a supplier considered but does not set it apart from the others being considered. In consumer markets, Romaniuk, Sharp and Ehrenberg reported in the Australasian Marketing Journal in 2007 that across 17 markets in Australia and the UK, an average of 11% of a brand's own users called it different and 10% called it unique. 77% of Apple's users did not call it either. The exception was Aldi, called different by 67% of its shoppers. Both are surveys of perception, and the second is not about business buyers.
What it means for a company selling complex technology.
A buyer comparing suppliers of batteries, sensors or quantum hardware is scanning a display. If every homepage in the category carries the same glow, the same promise and the same button, yours is the green T among green Xs and brown Ts: built from the same parts as everything around it. Turning the same look up louder does not change that. The experiment did not test loudness. It tested uniqueness, and the target found fast was the one with something nothing else had. Duncan and Humphreys add a useful point: search gets easier when the other items are alike. Read for a company, a category that all looks the same is the easiest place to be the different one. That reading is ours, not the paper's. The surveys add the condition. Only 14% of CEB's respondents saw a difference between suppliers worth paying for, so the look has to point at a difference that is real.
Strategy first, then the look.
Three moves, in this order. They are how we work, not a research finding. First, map the sameness: put every competitor's homepage side by side and note what they all say and show. Second, pick one difference: one thing you can prove that none of them claims. Third, give it one look: a colour, shape or image nobody else uses, then use it everywhere, on the site, the deck and the stand. The order is the point. A new colour on the same promise is just a louder black suitcase. Then test it. Put your homepage next to your three closest competitors. Cover every logo and every company name. Ask someone outside the company to point to yours. If they cannot, go back to the first move, not to the colour picker.
What to do with it.
THREE MOVES, IN THIS ORDERMap the sameness.
Put every competitor's homepage side by side. Note what they all say and show.
Pick one difference.
One thing you can prove that none of them claims. In CEB's 2009 survey, 14% of respondents saw a difference between suppliers worth paying for.
Give it one look.
A colour, shape or image nobody else uses. Then use it everywhere: the site, the deck and the stand.
ONE QUESTION
With every logo and company name covered, could someone outside your company pick your homepage out from your three closest competitors?
Yes or no. If it is no, start with the difference, not the colour.
Asked about this research.
NOT ANSWERED HERE?
tony@appart.giEVERY FIGURE HAS A SOURCE
Sources and limits.
- TREISMAN AND GELADE, 1980Treisman, A. M. and Gelade, G., 1980. A feature-integration theory of attention. Cognitive Psychology 12, 97 to 136. Experiment I, Table 1.
- DUNCAN AND HUMPHREYS, 1989Duncan, J. and Humphreys, G. W., 1989. Visual search and stimulus similarity. Psychological Review 96, 433 to 458.
- WOLFE, 1998Wolfe, J. M., 1998. What can 1 million trials tell us about visual search? Psychological Science 9(1), 33 to 39.
- CEB WITH GOOGLE, 2013CEB Marketing Leadership Council in partnership with Google, 2013. From Promotion to Emotion: Connecting B2B Customers to Brands. The 14% figure is credited there to the CEB 2009 Customer Experience Survey, n = 9,000.
- ROMANIUK AND COLLEAGUES, 2007Romaniuk, J., Sharp, B. and Ehrenberg, A., 2007. Evidence concerning the importance of perceived brand differentiation. Australasian Marketing Journal 15(2), 42 to 54.
- OUR METHODThe three moves and the logo test are APP/ART's own method, not research findings. The 9.3 times ratio, the times at 30 letters and Apple's 23% are our arithmetic on figures printed in the sources.
- LIMITSThe 1980 experiment is six people finding letters in a laboratory. It shows how the eye finds things, not how anyone decides to buy. The CEB figure is printed in a paper by CEB and Google, who both sell to marketers, and its sample is not described beyond n = 9,000. The 2007 study covers established consumer brands in two countries, and its sample sizes were not in the text we read. No study here tests the three moves.
Send your site. We name the first fix.
One line about where you are going is enough. The reply comes back in writing. No call needed.
Or book a 30 minute call
